1: %&latex
2: \documentclass[namedreferences,final]{SolarPhysics}
3:
4: \usepackage{graphicx}
5: \usepackage{url}
6: \usepackage[optionalrh]{spr-sola-addons}
7: \usepackage{rotating}
8:
9: \def\UrlFont{\sf}
10:
11:
12: \begin{document}
13: \begin{article}
14: \begin{opening}
15:
16:
17:
18: %###############################################################################
19: %#
20: %# OPENING OK
21: %#
22: %###############################################################################
23:
24: \title{Two-Dimensional Spectroscopy of Photospheric Shear Flows in a Small
25: $\delta$ Spot}
26:
27: \author{C.~\surname{Denker}$^{1,2}$\sep
28: N.~\surname{Deng}$^{2,3}$\sep
29: A.~\surname{Tritschler}$^{4}$\sep
30: V.~\surname{Yurchyshyn}$^{2,5}$}
31:
32: \runningauthor{Denker et al.}
33: \runningtitle{Two-Dimensional Spectroscopy of Photospheric Shear Flows}
34:
35: \institute{$^{1}$ Astrophysikalisches Institut Potsdam,\\
36: An der Sternwarte 16,
37: D-14482 Potsdam,
38: Germany,
39: email: \url{cdenker@aip.de}\\
40: $^{2}$ New Jersey Institute of Technology,
41: Center for Solar-Terrestrial Research,\\
42: 323 Martin Luther King Boulevard,
43: Newark, NJ~07102, U.S.A.\\
44: $^{3}$ California State University Northridge,
45: Department of Physics and Astronomy,\\
46: 18111 Nordhoff St,
47: Northridge, CA 91330, U.S.A.,
48: email: \url{na.deng@csun.edu}\\
49: $^{4}$ National Solar Observatory/Sacramento Peak,\\
50: P.O.~Box 62,
51: Sunspot, NM~88349, U.S.A.,
52: email: \url{ali@nso.edu}\\
53: $^{5}$ Big Bear Solar Observatory,\\
54: 40386 North Shore Lane,
55: Big Bear City, CA 92314, U.S.A.,
56: email: \url{vayur@bbso.njit.edu}}
57:
58: \date{Submitted: \ldots / Accepted: \ldots / Published online: \ldots}
59:
60:
61:
62: %##############################################################################
63: %
64: % ABSTRACT OK
65: %
66: %##############################################################################
67:
68: \begin{abstract}
69: In recent high-resolution observations of complex active regions, long-lasting
70: and well-defined regions of strong flows were identified in major flares and
71: associated with bright kernels of visible, near-infrared, and X-ray radiation.
72: These flows, which occurred in the proximity of the magnetic neutral line,
73: significantly contributed to the generation of magnetic shear. Signatures of
74: these shear flows are strongly curved penumbral filaments, which are almost
75: tangential to sunspot umbrae rather than exhibiting the typical radial
76: filamentary structure. Solar active region NOAA~10756 was a moderately complex,
77: $\beta\delta$ sunspot group, which provided an opportunity to extend previous
78: studies of such shear flows to quieter settings. We conclude that shear flows
79: are a common phenomenon in complex active regions and $\delta$ spots. However,
80: they are not necessarily a prerequisite condition for flaring. Indeed, in the
81: present observations, the photospheric shear flows along the magnetic neutral
82: line are not related to any change of the local magnetic shear. We present
83: high-resolution observations of NOAA~10756 obtained with the 65-cm vacuum
84: reflector at {\it Big Bear Solar Observatory} (BBSO). Time series of
85: speckle-reconstructed white-light images and two-dimensional spectroscopic data
86: were combined to study the temporal evolution of the three-dimensional vector
87: flow field in the $\beta\delta$ sunspot group. An hour-long data set of
88: consistent high quality was obtained, which had a cadence of better than
89: 30~seconds and sub-arcsecond spatial resolution.
90: \end{abstract}
91:
92: \keywords{Active Regions, Magnetic Fields;
93: Active Regions, Velocity Field;
94: Sun\-spots, Magnetic Fields;
95: Sunspots, Penumbra;
96: Flares, Pre-Flare Phenomena;
97: Flares, Relation to Magnetic Field}
98:
99: \end{opening}
100:
101:
102:
103: %##############################################################################
104: %
105: % INTRODUCTION OK
106: %
107: %##############################################################################
108:
109: \section{Introduction}
110:
111: The solar surface is highly dynamic and magnetic fields on various spatial
112: scales, from active regions and sunspots to network fields and faculae, are
113: intricately intertwined with the plasma motions. Photospheric flow fields are
114: important in the context of solar eruptive phenomena such as flares,
115: filament/prominence eruptions, and coronal mass ejections (CMEs). In
116: high-resolution studies of solar active region NOAA~10486, \inlinecite{Yang2004}
117: and \inlinecite{Deng2006} find strong shear flows close to the magnetic neutral
118: line, which are attributed to the slow energy build-up before its sudden release
119: in the flare. The local magnetic fields were highly non-potential and in this
120: case magnetic shear was the source of the free energy. However, the question
121: remains open if magnetic shear already existed when the magnetic field emerged.
122: This is of particular interest at the interface of solar dynamo theory and the
123: theory of flares and CMEs, since magnetic shear is related to highly-twisted
124: magnetic flux tubes, which can be generated by kink instability while the flux
125: tube rises from the bottom of the solar convection zone \cite{Fisher2000}.
126:
127: \inlinecite{Amari2000} showed that twisted flux tubes play a crucial role in the
128: theory of large-scale solar eruptive phenomena such as CMEs and two-ribbon
129: flares. If photospheric twist is applied to the footpoints of an arcade-like
130: magnetic field structure, the filament might become unstable and erupt. In this
131: scenario, $\delta$ spots play a distinct role as the origin of a complex
132: magnetic topology in the corona, which makes it conducive to eruptions.
133: Therefore, multi-polar fields are important in the study and modeling of flares,
134: filament/prominence eruptions, and CMEs. The evolution of flares can be
135: explained by a magnetic breakout process first described by
136: \inlinecite{Antiochos1999}, {\it i.e.}, magnetic reconnection at a null point
137: that is located high in the corona is opening initially low-lying sheared fields
138: leading to the flare, filament eruption and CME. \inlinecite{Aulanier2000}
139: further elaborate that a three-dimensional magnetic null point exists in the
140: corona above the $\delta$ spot and that reconnection occurs at the null point
141: long before the flare. However, there is an on-going debate centered on the
142: exact location and type of reconnection. For example, \inlinecite{Moore2001}
143: present an opposing view in terms of the standard bipolar model for eruptive
144: flares, where they conceptually tie the sheared and twisted core fields of a
145: bipolar active region to a runaway ``tether-cutting'' via impulsive
146: reconnection.
147:
148: In recent high-resolution observations, many phenomena were discovered, which
149: are directly related to eruptive phenomena such as the rapid penumbral decay as
150: a consequence of flares (\opencite{Wang2004}; \opencite{Deng2005};
151: \opencite{Liu2005}; \opencite{Sudol2005}), the detection of photospheric shear
152: flows near the magnetic neutral line and close to flare kernels
153: (\opencite{Yang2004}; \opencite{Deng2006}), and the first near-infrared (NIR)
154: continuum observations of flares \cite{Xu2004b}. Flare-associated shear flows
155: are not a new phenomenon. During the major flares in August 1972,
156: \inlinecite{Zirin1973} measured spectroscopically strong photospheric velocity
157: discontinuities of up to 6~km~s$^{-1}$ in the vicinity of neutral lines. In
158: more recent observations, converging Doppler velocities were observed in the
159: vicinity of $\delta$ spots. ({\it e.g.}, \inlinecite{Lites2002}). These flows
160: were interpreted as an interleaved system of field lines, where the outward
161: Evershed flow sharply bends downward and returns to the solar interior.
162:
163: In this study, we use two-dimensional spectroscopy to gain access to Doppler
164: velocities. Complementary observations of horizontal flows derived with local
165: correlation tracking (LCT, \opencite{November1988}) allow us to obtain a more
166: complete picture of the intricate flow field related to $\delta$ spots. These
167: high-cadence observations (30~seconds or less) of the photosphere in
168: combination with high-resolution observations of transition region and corona
169: have the potential to provide a comprehensive picture of rapid changes related
170: to magneto-convection and magnetic field evolution.
171:
172: In Section~2, we introduce the observations, provide detailed accounts of the
173: data calibration, and discuss their limitations in the context of our scientific
174: objectives. The results are presented in Section~3, where we describe the
175: intricate, three-dimensional flow fields encountered in NOAA~10756 with a
176: particular emphasis on the $\delta$ configuration. We discuss the results in
177: Section~4 and conclude this study of complex flow fields in sunspots with a
178: summary of the most important results in Section~5.
179:
180:
181:
182: %##############################################################################
183: %
184: % OBSERVATIONS OK
185: %
186: %##############################################################################
187:
188: \section{Observations}
189:
190: In the Summer of 2005, we obtained the first visible-light data with the 65-cm
191: vacuum reflector and its new high-order adaptive optics (AO) system.
192: Two-dimensional imaging spectroscopy and speckle masking imaging were employed
193: to measure, besides other quantities, solar vector flow fields.
194:
195: NOAA~10756 was selected as a promising target to study the dynamics in an active
196: region prior to an eruptive event, since it showed some moderate magnetic
197: complexity including a small $\delta$ configuration in its southern part.
198: However, the region did not produce any major M- or X-class flares during its
199: disk passage. In principle, scenarios with small flares (and even without any
200: flares) should receive the same attention as X-class events or geo-effective
201: CMEs. Otherwise, our views of such events might be skewed and we associate
202: phenomena with eruptive events that might be common to non-eruptive scenarios as
203: well. This is particularly true for high-resolution observations, which are rich
204: in detail and highly dynamic.
205:
206: At the beginning of its disk passage on 26~April 2005, NOAA~10756 produced
207: several C-class flares. The largest event was an impulsive C5 flare. On
208: 27~April, the region grew rapidly, predominantly in a $\delta$ spot attached to
209: the southern portion of the large sunspot. Despite its growth, NOAA~10756 only
210: produced lower-level C-class events. From 29~April to 1~May, the active region
211: reached its maximum area (about 1000 millionths of a solar hemisphere) before
212: it started to decay. We observed the moderately-complex $\delta$ configuration
213: on 2~May 2005, when it was located at heliographic coordinates
214: S8$^{\circ}$~W23$^{\circ}$ ($\mu = \cos\theta \approx 0.92$). At this time the
215: region was already decaying, however, it was still classified as a
216: $\beta\delta$ region. Only during the last few days of its disk passage a C7
217: and two M1 flares occurred, while the active region continued to decay.
218:
219: \begin{kaprotate}
220: \begin{table}
221: \caption{BBSO Observing Characteristics on 2 May 2005\label{TAB01}}
222: \begin{tabular}{lrllrrcll}
223: \hline
224: Type & $\lambda$ & Start & End & $\Delta t$ & $n$ & FOV & $s$ & $\delta t$\\
225: \hline
226: Speckle Reconstruction & $600 \pm 5$~nm & 17:01:07~UT & 17:30:57~UT & 30~s & 60 & $ 80^{\prime\prime} \times 80^{\prime\prime}$ & $0.08^{\prime\prime}$ pixel$^{-1}$ & 10~ms \\
227: Speckle Reconstruction$^{\rm a}$ & $600 \pm 5$~nm & 17:40:33~UT & 18:09:49~UT & 30~s & 60 & $ 80^{\prime\prime} \times 80^{\prime\prime}$ & $0.08^{\prime\prime}$ pixel$^{-1}$ & 10~ms \\
228: VIM & Fe\,{\sc i} 630.15~nm & 16:59:25~UT & 17:28:26~UT & 16~s & 105 & $ 80^{\prime\prime} \times 80^{\prime\prime}$ & $0.16^{\prime\prime}$ pixel$^{-1}$ & 80~ms \\
229: VIM & Fe\,{\sc i} 630.15~nm & 17:38:39~UT & 18:08:00~UT & 16~s & 110 & $ 80^{\prime\prime} \times 80^{\prime\prime}$ & $0.16^{\prime\prime}$ pixel$^{-1}$ & 80~ms \\
230: DVMG & Ca\,{\sc i} 610.30~nm & 15:53:10~UT & 22:30:40~UT & $\approx 80$~s & 220 & $288^{\prime\prime} \times 288^{\prime\prime}$ & $0.56^{\prime\prime}$ pixel$^{-1}$ & 30~ms$^{\rm b}$ \\
231: Filtergram & H$\alpha$ 656.30~nm & 15:54:01~UT & 22:31:31~UT & $\approx 80$~s & 248 & $300^{\prime\prime} \times 300^{\prime\prime}$ & $0.58^{\prime\prime}$ pixel$^{-1}$ & 60~ms \\
232: \hline
233: \multicolumn{9}{p{190mm}}{\rule{3ex}{0mm}NOTE.---The symbol $\lambda$ refers
234: to the wavelength, $\Delta t$ to the cadence, $n$ to the number of
235: observations in each time series, $s$ to the image scale, and $\delta t$ to
236: the exposure time. The speckle reconstructions and VIM data were taken with
237: the 65-cm vacuum reflector whereas the DVMG and filtergrams were obtained as
238: context data with the 25-cm refractor.}\\
239: \multicolumn{9}{p{190mm}}{\rule{3ex}{0mm}$^{\rm a}$The ten-minute gap between
240: the two high-resolution data sets occurred, since the optical trains had to
241: be re-aligned to compensate the pupil wobble on the AO wavefront sensor.}\\
242: \multicolumn{9}{p{190mm}}{\rule{3ex}{0mm}$^{\rm b}$The effective exposure time
243: of the magnetograms is actually larger ($\approx$ two~seconds), since 64 individual
244: exposures are combined to determine the degree of polarization Stokes
245: $V/I$.}
246: \end{tabular}
247: \end{table}
248: \end{kaprotate}
249:
250: The observations of NOAA~10756 are mainly based on observations with BBSO's
251: 65-cm vacuum reflector and 25-cm refractor. Table~\ref{TAB01} provides an
252: overview of the observing characteristics. Both telescopes have a suite of
253: instruments dedicated to active-region monitoring and space-weather research
254: \cite{Gallagher2002}. The observations with the 25-cm telescope provide the
255: context for the high-resolution observations. Two Apogee KX260 CCD cameras with
256: $512 \times 512$ pixel obtain filtergrams in the strong chromospheric
257: absorption lines Ca\,{\sc ii}\,K and H$\alpha$ with a one-minute cadence. The
258: image scale is about $0.58^{\prime\prime}$ pixel$^{-1}$ for both cameras
259: corresponding to a field-of-view (FOV) of about $300^{\prime\prime} \times
260: 300^{\prime\prime}$. The third optical bench on the 25-cm telescope is reserved
261: for the {\it Digital Vector Magnetograph} (DVMG, \opencite{Spirock2001}). A
262: quasi-continuum image and a line-of-sight (LOS) magnetogram are shown in
263: Figure~\ref{FIG01}. We used an intensity image and magnetogram of the {\it
264: Michelson Doppler Imager} (MDI, \opencite{Scherrer1995}) to establish the exact
265: image scale and coordinates for the DVMG data. Before aligning MDI and DVMG
266: data the solar position angle has to be taken into account. The image scale of
267: the DVMG is $0.56^{\prime\prime}$ pixel$^{-1}$ corresponding to a FOV of
268: $288^{\prime\prime} \times 288^{\prime\prime}$. The DVMG FOV is
269: $220^{\prime\prime} \times 220^{\prime\prime}$ after solar $P$-angle
270: correction. The $x$- and $y$-axes in Figure~\ref{FIG01} are given in disk
271: center coordinates and can be used as a reference for the high-resolution
272: observations. We usually depict the high-resolution images without rotation
273: correction. The DVMG line-wing image is commonly used as a reference, if we
274: have to co-align the H$\alpha$ filtergrams, the speckle reconstructions, and
275: two-dimensional spectroscopic data. Note that the line-wing images are used as
276: a substitute for continuum images.
277:
278: \begin{figure}[p]
279: %\centerline{\includegraphics[bb=0 0 406 800,height=179mm]{fig01.jpg}}
280: \centerline{\includegraphics[height=179mm]{fig01.eps}}
281: \caption{(a) Ca\,{\sc i} 610.3~nm line-wing filtergram and (b) longitudinal
282: magnetogram obtained with BBSO's DVMG system at 17:52~UT on 2~May 2005. The
283: magnetogram is displayed between $\pm 500$~G and the black (a) and white (b)
284: contours represent the magnetic polarity inversion line.}
285: \label{FIG01}
286: \end{figure}
287:
288: The high-resolution instrumentation of the 65-cm telescope is located in a
289: coud\'e laboratory. It consists of the high-order AO system, two imaging
290: spectro-polarimeters for observations in the visible and near infrared
291: wavelength regions (\opencite{Denker2003a}, \citeyear{Denker2003b}), and a fast
292: CCD camera system for image restoration \cite{Denker2005b}. The seeing
293: characteristics at BBSO were evaluated as part of the site survey for the {\it
294: Advanced Technology Solar Telescope} (ATST) and the site-specific results are
295: summarized in \inlinecite{Verdoni2007}. The high-order AO system
296: (\opencite{Rimmele2004a}; \opencite{Rimmele2004a}) is a collaboration between
297: BBSO and the {\it National Solar Observatory} (NSO). The optical set-up and
298: performance of the BBSO system are presented in \inlinecite{Denker2007b}. One
299: limitation of the AO system should be noted here, since it affects the data
300: collection in this study. A misalignment of the declination and right ascension
301: axes led to a wobble of the pupil image on the wavefront sensor. As a result,
302: the light path of the AO system and the post-focus instrumentation had to be
303: realigned every 30~minutes. This leads to a data gap of about five to ten
304: minutes between the 30-minute observing sequences.
305:
306: \begin{figure}[t]
307: %\centerline{\includegraphics[bb=0 0 800 800,width=80mm]{fig02.jpg}}
308: \centerline{\includegraphics[width=80mm]{fig02.eps}}
309: \caption{Schematic sketch of the small $\delta$ configuration in NOAA~10756.
310: Granulation, penumbrae, and umbrae are indicated as white, gray, and black
311: regions, respectively. The black line represents the magnetic neutral line.
312: The labels correspond to features discussed in the text.}
313: \label{FIG02}
314: \end{figure}
315:
316: Continuum images at $600 \pm 5$~nm of the North-East and South-East parts of
317: the active region were obtained from 17:01~UT to 17:31~UT and 17:40~UT to
318: 18:10~UT, respectively. The time series of speckle restored images has a
319: cadence of 30~seconds, {\it i.e.}, we restored a total of $2 \times 60$ images.
320: Each restored image is based on 100 short-exposure (ten~ms) images, which were
321: selected from 200 frames captured at 15~frames~s$^{-1}$. The FOV was about
322: $79.5^{\prime\prime} \times 79.5^{\prime\prime}$ with an image scale of
323: $0.078^{\prime\prime}$ pixel$^{-1}$. A target was inserted in a focal plane
324: following the tip-tilt mirror to accurately align the channels for image
325: restoration and two-dimensional spectroscopy. The image restoration procedure
326: has been described in detail in \inlinecite{Denker2005b}.
327:
328: In this study, we present for the first time scientific results obtained with a
329: new two-dimensional imaging spectrometer. In 2007, we will add polarization
330: optics to the spectrometer, which will enable us to measure the full Stokes
331: vector. The design of the imaging magnetograph systems at BBSO has been
332: outlined in \inlinecite{Denker2003a} and \inlinecite{Denker2003b}. As for now,
333: we use {\it Visible-light Imaging Magnetograph} (VIM) just as a spectrometer.
334: The imaging spectrometer was carefully tested, which includes a detailed
335: characterization of the Fabry-P\'erot etalon \cite{Denker2005a}. The
336: two-dimensional spectroscopic observations in the Fe\,{\sc i} 630.15~nm line
337: basically follow the continuum observations. However, they have a slightly
338: larger FOV of $84.4^{\prime\prime} \times 84.4^{\prime\prime}$. The CCD camera
339: was operated in a two $\times$ two pixel binning mode. Thus, the image scale
340: increased to $0.165^{\prime\prime}$ pixel$^{-1}$ compared to the continuum
341: observations. The cadence of the spectral line scans was 16~seconds and a total
342: of 214 scans were obtained. The filtergrams were taken at 91 wavelength points
343: in the Fe\,{\sc i} line separated by 1.2~pm corresponding to a scanned
344: wavelength interval of 0.11~nm. The exposure time of the filtergrams was 80~ms.
345:
346:
347:
348: %##############################################################################
349: %
350: % RESULTS
351: %
352: %##############################################################################
353:
354: \section{Results}
355:
356: Figures~\ref{FIG01} and \ref{FIG02} depicts the leading sunspot in active
357: region NOAA 10756 with predominantly negative polarity ({\it N}$_1$ and {\it
358: N}$_2$). The labels printed in italics refer to the schematic sketch shown in
359: Figure~\ref{FIG02}, which should serve as a guide to the many (small-scale)
360: features discussed in the following sections. Only a small $\delta$
361: configuration with opposite polarity ({\it P}$_1$) can be found in the southern
362: part of the region. The trailing positive polarity plage extends further to the
363: East -- well beyond the FOV shown in Figure~\ref{FIG01}. Flux concentrations
364: are limited to small pores or magnetic knots, which are invisible in MDI
365: magnetograms but sometimes recognizable in DVMG magnetograms at higher spatial
366: resolution. The leading part of NOAA~10756 consists of three distinct parts: a
367: dominant elliptical spot of negative polarity ({\it N}$_1$), a linear
368: arrangement (oriented along the North-South axis) of small sunspots with a
369: rudimentary penumbra ({\it N}$_2$) of the same polarity as the main sunspot,
370: and a small $\delta$ configuration ({\it P}$_1$). The North-South aligned spots
371: ({\it N}$_2$) are separated by a narrow granular region ({\it G}$_1$, vertical
372: box in Figure~\ref{FIG02}) from the main spot, while the penumbrae ({\it PU},
373: diagonal box in Figure~\ref{FIG02}) of the $\delta$ and main spot are fused
374: together. The magnetic polarity inversion line surrounding the negative flux of
375: the active region was derived from a low-pass filtered version of the
376: magnetogram. It is superimposed as black and white contours on the Ca\,{\sc
377: ii}\,K line-wing image and magnetogram shown in Figure~\ref{FIG01}. This
378: neutral line is also depicted in the high-resolution images.
379:
380: Since the FOV of the high-resolution observations is only $80^{\prime\prime}
381: \times 80^{\prime\prime}$, we decided to obtain two different data sets zooming
382: in on several small sunspots to the East ({\it N}$_2$) and a small $\delta$
383: configuration ({\it P}$_1$) to the south of the main sunspot ({\it N}$_1$).
384: Several small light-bridges branch out from the about $5^{\prime\prime}$-wide
385: channel of granulation ({\it G}$_1$), which separate individual umbral cores.
386: All umbral cores contain numerous umbral dots, which consist of peripheral
387: umbral dots with histories related to the migration of penumbral grains and a
388: background of fuzzier umbral dots. This linear array of sunspots formed a
389: well-organized rudimentary penumbra on its eastern side. The $\delta$ spot
390: ({\it P}$_1$) is located to the South-East of these small spots ({\it N}$_2$),
391: separated by a triangular shaped region of granulation. The main sunspot
392: connects to the remaining side of the triangle. Several tiny pores or magnetic
393: knots permeate this region of granulation ({\it G}$_2$, triangular box in
394: Figure~\ref{FIG02}), which is visible in both panels of Figure~\ref{FIG03}. The
395: $\delta$ spot consists of a single umbra containing several umbral dots, which
396: are, however, much fainter and less densely packed compared to the
397: aforementioned small sunspots. The umbral core is about $15^{\prime\prime}
398: \times 5^{\prime\prime}$. The penumbral filaments, which face away from the
399: main spot, point radially outward, whereas the filaments closest to the main
400: spot are strongly curved and wrapped around the umbral core of the $\delta$
401: spot. These almost-tangential filaments are the first visual indications of
402: highly twisted magnetic field lines.
403:
404: Tracing these individual penumbral filaments and penumbral grains in time
405: series of speckle reconstructions reveals a counter-clockwise twist. The
406: penumbral filaments of the main spot in the vicinity of the $\delta$ spot show
407: exactly the same twist leading to strong shear flows at the interface of the
408: two penumbrae. The penumbral segments, where penumbral filaments and grains
409: follow curved trajectories, are much wider than their counterparts with normal
410: radial motions. This might be a direct result of the Evershed flow. The
411: proximity of two penumbrae ({\it PU}) with colliding outflows leaves only two
412: possibilities: the subduction of one flow pattern or the horizontal deflection
413: of the opposing flows. Our observations of penumbral filaments, which are
414: tangential to the umbra of the $\delta$ spot well support mainly the latter
415: case. Strong shear flows are thus created, which will be aligned with the
416: polarity inversion line, if the two sunspots are of opposite polarity as in our
417: case. However, the subduction scenario put forward in \inlinecite{Lites2002}
418: would still apply to the locations where the dark penumbral filaments carrying
419: the Evershed flows terminate, {\it i.e.}, bend downward and return their
420: respective flows to the solar interior. Strands of penumbral filaments extend
421: for more than $25^{\prime\prime}$ while winding around the $\delta$ spot. These
422: strands show indications of a braided structure, which could be interpreted as
423: locally-confined subduction or horizontal twists in the flow channels
424: associated with the Evershed effect. Finally, the whirlpool-like magnetic
425: structure ({\it N}$_3$) near the western tip of the $\delta$ spot ({\it P}$_1$)
426: can serve as an example that localized shear flows and twist are present in
427: this active region even on scales of about $10^{\prime\prime}$.
428:
429: In Figure~\ref{FIG04}, we show dopplergrams obtained with the imaging
430: spectrometer. The Fe\,{\sc i} 630.15~nm line is broadened by the Zeeman effect.
431: This limitation should be kept in mind, when interpreting the findings based on
432: VIM data. The LOS velocity is determined by a Fourier-phase method
433: \cite{Schmidt1999}, which uses the entire line profile and is very insensitive
434: to noise. We follow the convention that redshifts are negative and blueshifts
435: are positive. Thus, dark areas in the dopplergrams shown in Figure~\ref{FIG04}
436: move away from the observer, while bright areas indicate motion towards the
437: observer. In the absence of an absolute wavelength calibration, we used quiet
438: Sun areas to establish a velocity calibration taking into account a convective
439: blueshift of 0.195~km~s$^{-1}$ \cite{Balthasar1988} for the Fe\,{\sc i}
440: 630.15~nm line at $\mu = \cos\theta = 0.92$. Since we are mostly interested in
441: the morphology and evolution of the flow fields, a rough interpretation of LOS
442: velocities as up- and downflows can be justified. In the following, we
443: concentrate on the major features of the LOS flows. A summary of the typical
444: flow velocities encountered in and around the active region are presented in
445: Table~\ref{TAB02}.
446:
447: \begin{table}[t]
448: \caption{Typical Velocities in Active Region NOAA~10756}
449: \begin{tabular}{lccccc}
450: \hline
451: Technique & Region & Quiet Sun & Penumbra & Umbra & Flow Kernel\\
452: \hline
453: LOS & 1 & $0.23 \pm 0.17$ & $0.44 \pm 0.27$ & $0.55 \pm 0.30$ & $\cdots$\\
454: LOS & 2 & $0.23 \pm 0.17$ & $0.39 \pm 0.26$ & $0.64 \pm 0.30$ & $0.96 \pm 0.15$\\
455: LCT & 1 & $0.46 \pm 0.27$ & $0.38 \pm 0.26$ & $0.26 \pm 0.14$ & $\cdots$\\
456: LCT & 2 & $0.49 \pm 0.28$ & $0.35 \pm 0.27$ & $0.28 \pm 0.16$ & $0.90 \pm 0.39$\\
457: \hline
458: \multicolumn{6}{p{100mm}}{\rule{3ex}{0mm}NOTE.---The region indices identify the
459: FOV in (1) the top panels and (2) the bottom panels shown in
460: Figures~\ref{FIG03} and \ref{FIG04}. LCT and LOS refer to the magnitude of
461: the velocities as measured with local correlation tracking and
462: two-dimensional spectroscopy. All velocities are given in units of
463: km~s$^{-1}$. The standard deviation of the velocities reflects the width of
464: the frequency distribution for the various solar regions and is not to be
465: confused with a measurement error.}
466: \end{tabular}
467: \label{TAB02}
468: \end{table}
469:
470: Some of the strongest LOS downflows are concentrated in the small pores and
471: magnetic knots contained in the triangle of granulation ({\it G}$_2$)
472: connecting the main spot ({\it N}$_1$), the linear group of sunspots ({\it
473: N}$_2$), and the $\delta$ spot ({\it P}$_1$). Typical downflow velocities in
474: these small-scale flux concentrations range from 0.6 to 1.0~km~s$^{-1}$. The
475: rudimentary penumbra belonging to the linear group of spots ({\it N}$_2$) has
476: the same velocity sign as the center-side penumbra of the main spot ({\it
477: N}$_1$).
478:
479: \begin{figure}[p]
480: %\centerline{\includegraphics[bb=0 0 406 800,height=164mm]{fig03.jpg}}
481: \centerline{\includegraphics[height=164mm]{fig03.eps}}
482: \caption{Speckle reconstructions of (a) the North-East (17:09~UT) and (b)
483: South-East (17:55~UT) parts of solar active region NOAA~10756 obtained on
484: 2~May 2005 with BBSO's 65-cm vacuum reflector. The thick black contours
485: represent the magnetic polarity inversion line. The thin black (white)
486: contours outline regions with LOS speeds of 1.0 and 1.2~km~s$^{-1}$ ($-1.0$
487: and $-0.8$~km~s$^{-1}$), respectively (see Figure~\ref{FIG04}). The black
488: arrows in the upper right corners indicate the direction toward disk
489: center. The $\delta$ spot is located in the center of the lower panel.}
490: \label{FIG03}
491: \end{figure}
492:
493: \begin{figure}[p]
494: %\centerline{\includegraphics[bb=0 0 406 800,height=164mm]{fig04.jpg}}
495: \centerline{\includegraphics[height=164mm]{fig04.eps}}
496: \caption{Dopplergrams displayed in the range of $\pm 1.5$~km~s$^{-1}$ of (a) the
497: North-East (17:09~UT) and (b) South-East (17:55~UT) parts of solar active
498: region NOAA~10756 obtained on 2~May 2005 with BBSO's 65-cm vacuum reflector.
499: The thick black contours represent the magnetic polarity inversion line.
500: The thin white (black) contours outline regions with LOS speeds of 1.0 and
501: 1.2~km~s$^{-1}$ ($-1.0$ and $-0.8$~km~s$^{-1}$), respectively. The white
502: arrows in the upper right corners indicate the direction toward disk center.
503: The dopplergrams were slightly cropped to fit the FOV of the continuum
504: speckle reconstructions.}
505: \label{FIG04}
506: \end{figure}
507:
508: The most conspicuous LOS velocity feature of the $\delta$ spot is its penumbra.
509: The line separating upflows and downflows (contained in the diagonal box {\it
510: PU} in Figure~\ref{FIG02}) basically follows the magnetic polarity inversion
511: line. However, it is displaced by about $5^{\prime\prime}$. A closer inspection
512: of the magnetograms reveals that the positive polarity penumbra of the $\delta$
513: spot contains several elongated inclusions of of small-scale negative polarity
514: features, {\it i.e.}, without smoothing several localized magnetic neutral
515: lines would exist. The neutral line shown in Figures~\ref{FIG01} to \ref{FIG06}
516: was derived from a smoothed magnetogram, which eliminate many of these small
517: ``polarity islands'' and thus effectively decreases the majority flux
518: surrounding these locations. As a result, the magnetic neutral line is shifted
519: towards the $\delta$ spot. However, the separation line between up- and
520: downflows coincides well with the strongest magnetic field gradient. These
521: observations support the scenario in which the penumbrae of the $\delta$ spot
522: and the main spot behave primarily as separate entities, {\it i.e.}, the
523: Evershed flow is deflected. However, intrusions of opposite polarity flux might
524: occur near this interface, which appear as the aforementioned braided structure
525: of penumbral filaments. Localized subduction as suggested by
526: \inlinecite{Lites2002} might play a role in these cases. Only high-resolution
527: two-dimensional spectro-polarimetry with sub-arcsecond resolution can answer
528: this question.
529:
530: Since the penumbral filaments on the western side are almost tangential, they
531: behave essentially as if they belong to the limb-side penumbra. Strong flow
532: kernels are embedded in both colliding penumbrae. The size of these elongated
533: flow kernels is about $5^{\prime\prime} \times 2^{\prime\prime}$ and they are
534: aligned with the local penumbral filaments. Flows in these kernels are
535: typically larger than 0.8~km~s$^{-1}$ and reach almost 1.8~km~s$^{-1}$. No dark
536: features in the speckle restored continuum images (see Figure~\ref{FIG03}) are
537: associated with these flow kernels. Analyzing the time series of dopplergrams,
538: we find that the flow kernels are slowly evolving, long-lived features (at
539: least 30~minutes) that do not partake in the penumbral proper motions seen in
540: time series of speckle restored continuum images. Since no magnetograms with a
541: spatial resolution comparable to that of the two-dimensional spectroscopic data
542: or the speckle restored images was available, the questions remains open, if
543: the flow kernels are associated with intrusion of opposite polarity flux into
544: the the penumbral shear flow areas.
545:
546: In Figures~\ref{FIG05} and \ref{FIG06}, we show high-resolution flow maps based
547: on time series of speckle reconstructed images. The two figures depict the
548: detailed proper motions in the northern and southern parts of active region
549: NOAA 10756, respectively. Averaging 60 individual flow maps over a period of
550: about 30~minutes only leaves persistent flow patterns. The smallest reliable
551: features in the flow maps have a size of about $1^{\prime\prime}$. For this
552: reason, we opted to use a color representation to indicate flow direction and
553: magnitude rather than relying on drawing vector maps with inferior resolution.
554: A 12-sector RGB color wheel has been added as a visualization guide in the
555: upper-right corners of the azimuth maps. In addition, contour lines of the
556: granulation/penumbra boundary (white) and the penumbra/umbra (gray) were
557: superimposed on the flow maps to ease the interpretation of the highly detailed
558: maps. The areas with granulation are characterized by small-scale patches
559: (having the size of individual granular cells) with large variations of the
560: azimuth angle. In addition, in regions with small-scale magnetic fields, {\it
561: i.e.}, brightenings in the line core intensity maps, we find some of the
562: largest proper motions. Typical horizontal velocities related to Fe\,{\sc i}
563: line gaps frequently reach 1.25~km~s$^{-1}$, whereas velocities in areas of
564: normal granulation are typically around 0.75~km~s$^{-1}$. A comparison of the
565: typical LCT and LOS velocity values is given in Table~\ref{TAB02} for different
566: regions and features in the active region.
567:
568: The proper motions in penumbral regions retain the elongated character of the
569: penumbral filaments. The largest velocities are generally found near the
570: granulation/penumbra boundary, {\it i.e.} they are encountered in the vicinity
571: of the white contour lines in Figure~\ref{FIG05}b. Conversely, the lowest
572: velocities are found in the umbral regions of the sunspots. However, this might
573: be related to the low signal-to-noise ratio in these dark regions, where only
574: the faint umbral dots contribute to the correlation signal measured with the
575: LCT technique. The high LOS velocities, which were associated with the small
576: pores and magnetic knots in the previously mentioned triangular granular region
577: ({\it G}$_2$), have no counterparts in the LCT maps. The most prominent feature
578: in the horizontal flow maps is located in the center of Figure~\ref{FIG06}b as
579: indicated by a white box. Two flow kernels, one elongated and the other with
580: circular shape, are clearly visible. The diameter of the circular kernel is
581: about $5^{\prime\prime}$ and the other kernel is about $3^{\prime\prime}$ wide
582: and stretches out to about $12^{\prime\prime}$. The velocity in these flow
583: kernels approaches 2.5~km~s$^{-1}$. In addition, they are co-aligned with flow
584: kernels in the LOS velocity maps. The two flow kernels are within
585: $5^{\prime\prime}$ of each other and separated by a line tracing the strongest
586: magnetic field gradient, which parallels the magnetic polarity inversion line
587: about $5^{\prime\prime}$ to the North-West. Compared to previous observations
588: of active region NOAA~10486 (see \opencite{Yang2004}, especially Figures~2b and
589: 2c), the flow interface is harder to trace in the azimuth and magnitude maps,
590: since it does not exactly follow the magnetic polarity inversion line. However,
591: the proximity of green and purple colors $5^{\prime\prime}$ to the North-East
592: of the magnetic neutral lines indicates flows in opposite direction. Thus,
593: strong shear flows are present in NOAA~10756 as well.
594:
595: \begin{figure}[p]
596: %\centerline{\includegraphics[bb=0 0 427 680,height=172mm]{fig05.jpg}}
597: \centerline{\includegraphics[height=172mm]{fig05.eps}}
598: \caption{Horizontal flow fields in NOAA~10756 on 2~May 2005. LCT was applied to
599: sequences of 60 speckle reconstructions obtained during the time period from
600: 17:01~UT to 17:31~UT (northern part). (a) The color-coded direction of the
601: optical flows (see the ``compass rose'' in the upper right corner) and (b)
602: the corresponding horizontal flow speeds. The white arrow indicates the
603: direction toward disk center. The FOV is slightly smaller compared to
604: Figure~\ref{FIG03} due to image rotation in the coud\'e laboratory.}
605: \label{FIG05}
606: \end{figure}
607:
608:
609: \begin{figure}[p]
610: %\centerline{\includegraphics[bb=0 0 427 680,height=172mm]{fig06.jpg}}
611: \centerline{\includegraphics[height=172mm]{fig06.eps}}
612: \caption{Horizontal flow fields in NOAA~10756 on 2~May 2005. LCT was applied to
613: sequences of 60 speckle reconstructions obtained during the time period from
614: 17:40~UT to 18:10~UT (southern part). (a) The color-coded direction of the
615: optical flows (see the ``compass rose'' in the upper right corner) and (b)
616: the corresponding horizontal flow speeds. The white arrow indicates the
617: direction toward disk center. The white box indicates the region with
618: strong horizontal shear flows.}
619: \label{FIG06}
620: \end{figure}
621:
622: We define magnetic shear according to \inlinecite{Wang1994} as the angular
623: difference between the measured and calculated potential transverse field. In
624: Figure~\ref{FIG07}, we show in the background a longitudinal magnetogram of the
625: small $\delta$ configuration taken at about the same time as the data shown in
626: Figure~\ref{FIG01}. The FOV of the vector magnetogram is $50^{\prime\prime}
627: \times 50^{\prime\prime}$. The orange arrows indicate magnitude and direction
628: of the measured transverse field component. The $180^{\circ}$ ambiguity was
629: resolved by applying the acute angle method, which requires that the transverse
630: components of the observed and modeled (in this case potential) fields make an
631: cute angle. Recent studies in the framework of the SDO/HMI--CSAC Azimuth
632: Ambiguity Resolution Workshop \cite{Metcalf2006} showed that this method can
633: correctly resolve ambiguity for about 90\% of strong transverse magnetic
634: fields. The magnetic shear angle has been computed for a $2^{\prime\prime}$
635: neighborhood along both sides of the magnetic neutral line (shown as a thick
636: white line). The magnetic shear angle in the interval $[-90^{\circ},
637: +90^{\circ}]$ is mapped to a rainbow color-scale shown at the right side of
638: Figure~\ref{FIG07}. Two areas with distinct magnetic shear angles exist along
639: the magnetic neutral line. One area corresponds to the diagonal box {\it PU} in
640: Figure~\ref{FIG02}, where the photospheric shear flows are observed. The mean
641: magnetic shear angle is negative (blue) in this region. In contrast, the
642: magnetic shear angle along the magnetic neutral line immediately to the south
643: of is positive (red).
644:
645: Since we have 243 vector magnetograms during the time period from 16:43~UT to
646: 22:32~UT, we can study the temporal evolution of the magnetic shear and relate
647: it to the observed photospheric shear flows. For this purpose, we defined a
648: neighborhood of $2^{\prime\prime}$ along both sides of the magnetic neutral
649: line as indicated in Figure~\ref{FIG07} but now for all vector magnetograms.
650: The average magnetic shear is computed for the entire neighborhood area of the
651: neutral line visible in the $50^{\prime\prime} \times 50^{\prime\prime}$ FOV of
652: Figure~\ref{FIG07}. Its temporal evolution is presented by the black curve in
653: Figure~\ref{FIG08}. The magnetic shear increased from about $+9^{\circ}$ at
654: 17:00~UT to about $+21^{\circ}$ at 22:30~UT. However, this increase occurs
655: predominantly in the southern part of the neutral line with positive magnetoc
656: shear angle, whereas in proximity to the region with photospheric shear flows,
657: the magnetic shear angle remains fairly constant at about $-36^{\circ}$. Shear
658: angle values around $\pm 30^{\circ}$ are also commonly encountered in
659: non-$\delta$ areas of sunspots. In $\delta$ spots with major flares, the
660: magnetic shear angle is often close to $\pm 90^{\circ}$ ({\it e.g.},
661: \opencite{Hagyard1984}).
662:
663: In addition to the photospheric observations, we obtained chromospheric data,
664: which provides some clues to the overall magnetic field topology. Three
665: distinct loop system indicated by white boxes in Figure~\ref{FIG09} can be
666: identified in the H$\alpha$ filtergrams: i) small loops connecting the main
667: spot with opposite polarity flux in the sunspot moat in the north-western part
668: of the active region, ii) large-scale loops associated with the linear group of
669: small sunspots ({\it N}$_2$) in the North-East, which terminate in a region of
670: extended plage, and iii) a large loop between the $\delta$ spot ({\it P}$_1$)
671: and the umbra of the main sunspot ({\it N}$_1$). All loop systems have in
672: common that they point away from the center of the main spot and leave the
673: central region of the sunspot group devoid of any H$\alpha$ loops. In
674: particular, no active-region filament formed along the polarity-inversion line.
675: In time series of H$\alpha$ filtergrams, we find some indications of activity
676: related to the loop systems, which is most prominent in the flux system towards
677: the North-East (Box~1 in Figure~\ref{FIG09}). Here, we find frequent
678: brightenings of the H$\alpha$ plage, which originates at the footpoint of a
679: small filament routed in the linear sunspot group. The $\delta$ spot filament
680: (Box~3) is fairly inactive with the exception of one short-term brightening
681: around 18:07~UT, which traveled at high speeds from the umbra of the main spot
682: to approximately the apex of the H$\alpha$ loop. Even though the connection of
683: the $\delta$ spot filament to the main sunspot might not be clear in
684: Figure~\ref{FIG09}, the trajectory of this brightening undoubtedly confirms
685: this connection. This feature is very similar to transient flow fields in
686: filaments, which were observed by \inlinecite{Chae2000} in high-resolution
687: times-series of H$\alpha$ filtergrams. In addition to these transient flows, we
688: find indications of twisted flows near the loop top. These twisted flows are
689: especially prominent during times when the absorption is increasing at the apex
690: of the loop.
691:
692: \begin{figure}[t]
693: %\centerline{\includegraphics[bb=0 0 660 494,width=120mm]{fig07.jpg}}
694: \centerline{\includegraphics[width=120mm]{fig07.eps}}
695: \caption{Vector magnetogram of the small $\delta$ configuration. The
696: transverse magnetic field is given as orange vectors superposed on a
697: gray-scale representation of the longitudinal magnetogram, which is scaled
698: between $\pm 500$~G. The magnetic neutral line is shown in white. The
699: magnetic shear angle is displayed in a neighborhood of $\pm 2^{\prime\prime}$
700: along the magnetic neutral line. The magnetic shear angle is encoded
701: according to the color index on the right.}
702: \label{FIG07}
703: \end{figure}
704:
705: We used data of the {\it Transition Region and Coronal Explorer} (TRACE,
706: \opencite{Handy1999}) to compare the chromospheric with the coronal loop
707: structure above NOAA~10756. Figure~\ref{FIG10} is a superposition of a Ca\,{\sc
708: i} 610.3~nm line-wing filtergram and a negative TRACE 171~\AA\ image, which
709: maps coronal loop structures with temperatures of about $1 \times 10^{6}$~K.
710: Note that TRACE observations of NOAA~10756 were not available during the BBSO
711: observations. Therefore, we selected the first available image on the
712: following day. In general, the chromospheric and the coronal loop topology is
713: very similar. Again, the absence of loop structures (Box~1 in
714: Figure~\ref{FIG09}) above the sunspots is worth noticing. As expected, the
715: low-lying loop system related to the sunspot moat is invisible in the TRACE
716: 171~\AA\ image. However, the other two loop systems are clearly visible. The
717: loop system in the North-East, which is associated with the extensive plage
718: region, is routed in the linear group of sunspots ({\it N}$_2$) and the main
719: sunspot ({\it N}$_1$). On the other side of NOAA~10756, the loop system related
720: to the $\delta$ configuration ({\it P}$_1$) is much more pronounced compared to
721: the chromospheric observation. Several individual loops are contained in this
722: loop system.
723:
724: \begin{figure}[t]
725: %\centerline{\includegraphics[bb=0 0 800 599,width=100mm]{fig08.jpg}}
726: \centerline{\includegraphics[width=100mm]{fig08.eps}}
727: \caption{Temporal evolution of the mean magnetic shear angle along the
728: entire magnetic neutral line shown in Figure~\ref{FIG07} (black curve). The
729: temporal evolution of the mean magnetic shear angle for region {\it PU} with
730: photospheric shear flows (negative shear) and for the region immediately to
731: the South (positive shear) is given by the two gray curves at the bottom and
732: top, respectively.}
733: \label{FIG08}
734: \end{figure}
735:
736:
737:
738: %##############################################################################
739: %
740: % DISCUSSION
741: %
742: %##############################################################################
743:
744: \section{Discussion}
745:
746: The most striking features in LCT and LOS velocity maps are localized areas of
747: strong flows. They fall into two categories: either they are associated with
748: small sunspots and pores embedded in a triangular region of granulation ({\it
749: G}$_2$), which is enclosed in the sunspot group, or they are located on
750: opposite sides of the magnetic polarity inversion line in regions of strong
751: shear flows (diagonal box {\it PU} in Figure~\ref{FIG02}). In the first case,
752: the observed downflows are typical for small sunspot and pores. Typically, the
753: velocity field of pores is close to zero near their center with strong
754: downflows in their periphery ({\it e.g.}, \opencite{Leka1998};
755: \opencite{Sankarasubramanian2003}).
756:
757: \begin{figure}[t]
758: %\centerline{\includegraphics[bb=0 0 740 720,height=120mm]{fig09.jpg}}
759: \centerline{\includegraphics[height=120mm]{fig09.eps}}
760: \caption{H$\alpha$ filtergram of NOAA~10756 obtained at 17:51~UT on 2~May 2005
761: with BBSO's 25-cm refractor. The white boxes mark three H$\alpha$ loop
762: systems discussed in the text.}
763: \label{FIG09}
764: \end{figure}
765:
766: The case of the flow kernels near the magnetic neutral line is different. These
767: kernels are not associated with dark features but with the sheared penumbral
768: velocity field. \inlinecite{MartinezPillet1994} reported strong, isolated
769: downflows near the neutral line of a $\delta$ spot, which were as large as
770: 14~km~s$^{-1}$. The downward motions were accompanied by upward flows of about
771: 1.5~km~s$^{-1}$ in a nearby region. These velocities were measured with the
772: {\it Advanced Stokes Polarimeter} (ASP) at NSO/Sacramento Peak by determining
773: the zero-crossing of Stokes $V$ profiles. This velocity signal might not be as
774: easily visible in the Doppler shift of the Stokes $I$ signal. Since our
775: spectral resolution is inferior to the spectrograph-type ASP instrument and we
776: have to rely on Stokes $I$ line profiles, the velocity signal of the flow
777: kernels in our study is not expected to be as high. The strongest upflow was
778: encountered in the disk-side penumbra of the main sunspot, which is accompanied
779: by kernel with smaller downflows on the other side of the polarity inversion
780: line. Despite the discrepancy in the velocity signal both observations of flow
781: kernels show a variety of commonalities. The flow kernels appear as pairs of
782: opposite flows on both sides of the strongest magnetic field gradient between
783: the penumbrae of the $\delta$ spot and main spot. They are long-lived features
784: (30~minutes to 3~hours). Typical sizes are in the range from $2^{\prime\prime}$
785: to $5^{\prime\prime}$. According to \inlinecite{MartinezPillet1994}, non-linear
786: phenomena in the transition region and chromosphere are responsible for the
787: observed high speeds. The strongly-curved penumbral filament in the limb-side
788: penumbra of the $\delta$ spot, which are almost tangential to the umbra,
789: indicate the presence of strong magnetic twist. \inlinecite{Leka2005} showed
790: that the onset of kink instability requires sufficient magnetic twist (flux
791: tubes with more than $2\pi$ winds) in the solar active region. At this moment,
792: it is not clear if two-dimensional spectro-polarimetry can provide enough
793: information to quantify the amount of twist. However, it can identify emerging
794: helicity-carrying flux tubes and help to identify areas, which are susceptible
795: to kink-instability. These quantities are valuable for any operational space
796: weather monitoring and predicting system.
797:
798: \begin{figure}[t]
799: %\centerline{\includegraphics[bb=0 0 800 800,height=120mm]{fig10.jpg}}
800: \centerline{\includegraphics[height=120mm]{fig10.eps}}
801: \caption{Superposition of a gray-scale Ca\,{\sc i} 610.3~nm line-wing
802: filtergram observed at 17:52~UT on 2~May 2005 and a co-aligned, false-color,
803: negative 171~\AA\ image obtained with TRACE at 03:48~UT on 3~May 2005
804: showing the coronal morphology of active region NOAA~10756 at about
805: $1 \times 10^6$~K.}
806: \label{FIG10}
807: \end{figure}
808:
809: $\delta$ spots have been identified as the most flare-productive active regions
810: \cite{Zirin1987}. Why does then the $\delta$ configuration in NOAA~19756
811: produce so little activity? For one, the $\delta$ configuration is not a very
812: close one, {\it i.e.}, it is not an ``island configuration''. Indeed, one could
813: argue that the $\delta$ spot is not located in the penumbra of the main spot
814: but possesses its own penumbra, which is just in very close proximity to the
815: main spot's penumbra. The penumbrae seem to merge in low-resolution
816: observations. However, the high-resolution observations clearly reveal the
817: distinct identity of both the $\delta$ and main spot. This observation is
818: confirmed by the LOS and horizontal flow patterns associated with the two
819: penumbrae. \inlinecite{Sammis2000} have provided a comprehensive study of the
820: relationship between flares and $\delta$ spots. $\beta\gamma\delta$ regions are
821: the most prolific flaring regions producing many more large flares compared to
822: other regions of similar size. $\beta\delta$ spots, such as NOAA~10756, are
823: ranked (see Figure~2 in \opencite{Sammis2000}) just below $\beta\gamma\delta$
824: spots on a par with $\delta$ spots. We would expect predominantly M-class flares
825: with some C-class and occasional X-class flare from a $\beta\delta$ region.
826: Considering the size of NOAA~10756 (about 1000 millionths of a solar
827: hemisphere), the likelihood for producing M-class flares was given.
828:
829: In the region of colliding penumbrae (diagonal box {\it PU} in
830: Figure~\ref{FIG02}), we observe a negative (left-handed) magnetic shear along
831: the neutral line (Figure~\ref{FIG07}). The observed photospheric shear flow,
832: which is carried by the more horizontal Evershed flow channels, would tend to
833: decrease the magnetic shear, thus leading to a more potential configuration of
834: the local magnetic field. The temporal evolution of the magnetic shear in
835: region {\it PU} shows a fairly constant negative shear angle of about
836: $-36^{\circ}$, whereas in other regions of the neutral line the magnetic shear
837: increases (see Figure~\ref{FIG08}). The gradual build-up of magnetic shear in
838: these regions can be attributed to motions of the small $\delta$ configuration
839: and main sunspot on larger scales. The lack of major flare activity could,
840: therefore, also have its explanation in the fact that locally, the photospheric
841: shear flows might inhibit an increase of magnetic shear. At least it remained
842: constant at a level, which is common for non-$\delta$ regions in sunspots
843: \cite{Hagyard1984}. This situation is very different from the instantaneous
844: magnetic shear evolution observed during major flares \cite{Wang1994} and
845: differs from recent observations of photospheric shear flows in flaring active
846: region NOAA~10486 (\opencite{Yang2004}; \opencite{Deng2006}).
847:
848: Based on our observations, we conclude that the polarity inversion line was too
849: short and sufficient shear was not available to create a channel enabling the
850: formation of an active region filament. The chromospheric and coronal loops are
851: pointing in two opposite directions leaving the volume above the sunspot group
852: almost void of any loops. Therefore, the magnetic field topology is much
853: different from the ``spine'' fields and their associated ``fan'' surface
854: surrounding a parasitic polarity, which have been observed in many flares with
855: filament eruptions and accompanying CMEs, {\it e.g.},
856: \inlinecite{Aulanier2000}, who studied the ``Bastille Day'' flare on 1998
857: July~14.
858:
859:
860:
861: %##############################################################################
862: %
863: % CONCLUSIONS
864: %
865: %##############################################################################
866:
867: \section{Conclusions}
868:
869: We have presented the first science observations with newly developed
870: visible-light post-focus instruments at BBSO. The combination of a high-order AO
871: system with an imaging spectrometer and a fast CCD camera system for speckle
872: observations enabled us to measure the three-dimensional flow fields in and
873: around solar active region NOAA~10756.
874:
875: We found well-defined flow kernels embedded in a $\delta$ configuration, where
876: both the horizontal and LOS velocities reached up to 1.8~km~s$^{-1}$ and 2.5~km
877: s$^{-1}$, respectively. These flow kernels, which are one signature of the
878: atypical flow structure in $\delta$ spots, were previously associated with
879: flares ({\it e.g.}, \opencite{Yang2004}; \opencite{Deng2006}). However, since
880: these flow kernels are present in this relatively quiet active region, we
881: conclude that they are common features in magnetically complex active regions.
882: The observed photospheric flows are related to the magnetic fields of the more
883: horizontal Evershed flow channels. A connection to the more vertical fields of
884: the uncombed penumbra still needs to be established, since these are the fields
885: straddling the magnetic neutral line. However, this task requires future
886: observations with diffraction-limited spectro-polarimetry with good spectral
887: resolution. However, the well-established flow interface between the opposite
888: polarity penumbrae in NOAA~10756 did not have a counterpart in the respective
889: magnetic flux systems. Here, the situation is more complex. Small-scale
890: intrusions of opposite polarity flux into the $\delta$ spot penumbra might
891: locally increase the flow speed, while on the other side, locally contribute to
892: the dissipation of magnetic flux. We find that the magnetic shear remains
893: fairly constant in the regions with the photospheric shear flows and that the
894: direction of the flow might locally decrease magnetic shear. This scenario
895: hints at an intricate interaction between the global flow field established by
896: two deflected penumbrae and localized phenomena in penumbrae such as flow
897: kernels and inclusions of opposite polarity features. In any case, the peculiar
898: flows in the vicinity of $\delta$ spots are indicators for rapid field
899: evolution, flux emergence, and the presence of twisted fields.
900:
901: Comparing our high-resolution observations with active region NOAA~10486
902: (\opencite{Xu2004b}; \opencite{Deng2006}), we find that the polarity inversion
903: line separating the $\delta$ spot from the main spot is much shorter. Only a
904: single location of strong shear flows was identified. Two intrusions of
905: negative polarity flux into areas with opposite-polarity plage limit the region
906: along the polarity inversion line where shear could be built-up. Thus, the
907: short magnetic neutral line is not conducive to the formation of an active
908: region filament. In addition, the overlying coronal and chromospheric field
909: topology as deduced from TRACE 171~\AA\ and BBSO H$\alpha$ filtergrams is much
910: different from the non-potential field above a filament channel, {\it i.e.} two
911: loop systems are deflected in opposite direction pointing away from the central
912: part of NOAA~10756. Therefore, magnetic reconnection above the polarity
913: inversion line is highly unlikely.
914:
915: In summary, observations with high temporal, spatial, and spectral resolution
916: as well as with sufficient magnetic sensitivity are vital in advancing our
917: understanding of solar eruptive phenomena and how they affect Earth and the
918: near-Earth environment.
919:
920:
921:
922: %##############################################################################
923: %
924: % ACKNOWLEDGMENTS
925: %
926: %##############################################################################
927:
928: \acknowledgements We would like to thank Wolfgang Schmidt and an anonymous
929: referee, whose careful reading of the manuscript led to substantial
930: improvements. This work was supported by NSF under grant ATM 00-86999, ATM
931: 02-36945, IIS~ITR 03-24816 and AST~MRI 00-79482 and by NASA under grant NAG
932: 5-12782. The National Solar Observatory is operated by the Association of
933: Universities for Research in Astronomy under a cooperative agreement with the
934: National Science Foundation, for the benefit of the astronomical community. The
935: data analysis including the graphics was performed in the {\it Interactive Data
936: Language} (IDL) by ITT Visual Information Solutions.
937:
938:
939:
940: %##############################################################################
941: %
942: % BIBLIOGRAPHY
943: %
944: %##############################################################################
945:
946: \bibliographystyle{spr-mp-sola}
947: \bibliography{spr-mp-sola-jour,cdenker}
948: %\input{ms.bbl}
949:
950:
951:
952: \end{article}
953: \end{document}
954: